Still a Long Wait for Microsoft Longhorn

You’ll see a lot of stories about Microsoft’s next-generation of Windows, code-named Longhorn, popping up during the next couple of days. Jim Allchin, Microsoft’s group vice president for platforms, has been on the road with a recent build of the OS and an updated message designed to get The Media (mea culpa) chatting anew after a period of relative quiet. Oh, and in the interest of making even more noise, HAPPY 20th BIRTHDAY TO WINDOWS!! (Microsoft will be repeating that theme as well.)

In a nutshell, here’s what you need to know about Longhorn.

1) It’s a long, long ways off. Microsoft is sticking to its claims of a 2006 ship date for the client version, but they’re describing it as “Holiday 2006,” which means they could declare victory over slipped schedules even if CompUSA employees are stocking shelves with boxes covered in shrinkwrap so fresh it burns their fingers as the strains of Auld Lang Syne are coming from the TV in the employee breakroom.

And the server version will ship sometime in 2007. That’s two years, which is great news for developers who have to get up to speed on the new Longhorn development tools and then need to update their applications in time to piggyback on the millions of marketing dollars Microsoft will undoubtedly spend on the launch.

As for me, I’m going to celebrate a couple of birthdays myself before I start making any upgrade plans.

2) Longhorn is slick: Have no doubt. The user interface positively glitters, with cool fade effects, transparent title bars and the like. This is all thanks the the much-hyped Avalon interface engine, otherwise known as “The Graphics Card Vendor’s Best Friend” because of the hardware you’ll need to take full advantage of the system. (Microsoft was demoing Longhorn on one of those shiny red Ferrari-themed Acer notebooks, complete with an AMD Athlon 64-bit processor. I’m always more impressed when people demo on something more realistic, though given the timeline, you may not be able to buy a 32-bit machine other than used on eBay by the time Longhorn arrives)

And the search features I saw put XP’s file-finding functions to shame. Forget about folders, now you get customizable views. Just start typing and documents begin to appear. Add keywords. Categorize. Group by creator. By Date. Heck, sort ’em by color, scent and favorite movie. Search through metadata that Windows will automatically generate based on content. The search feature alone would be worth the upgrade—if Microsoft were releasing Longhorn now. But the competition—including X1, Google, and Yahoo—have a couple of years to build all these features and more into their own products, which will undoubtedly negate some if not all of Microsoft’s advantage.

3) Longhorn will be all about extensibility. XML runs rampant throughout the platform, and developers will supposedly be able to use Longhorn’s openness to quickly enhance existing features and build new ones. Only time will tell exactly how “open” Microsoft ultimately will be, but at least it appears to be taking a couple of steps in the right direction. I’m not holding my breath in expectation that Redmond will break its eternal tradition of “open…but”—though we can always hope.

4) It’ll be the most secure Windows yet. Somehow I think IT managers are going to get a little tired of having this drum banged in their ears from here to eternity (shouldn’t the operating systems they’re buying now be secure?), but get used to it. With Longhorn, Microsoft will change some core functionality in Windows’ security model. In XP for instance, the vast majority of users run as administrators. Why? Because XP pretty much divides all users into one of two groups: “Administrator” and “Virus” (aka, “User”). (Heck, some viruses actually have more privileges in XP than do standard “Users”—thus the need for XP SP2).

In Longhorn, IT admins will be allowed to get much more granular. They’ll be able to restrict everyday users from seriously dangerous activities while still letting them get their jobs done, say by enabling a wireless interface when they need to access their e-mail from a hotel room without it involving 30 minutes with the help desk. Individual applications will also be able to get their own security level. So an administrator could be doing disk management on a system while her browser window hums away at a lower level, hopefully eliminating—or at least making more difficult—a number of now-common attacks.

5) Longhorn will be the foundation for the next ten years of Windows. This goes a long way in explaining why Microsoft is taking such pains—and causing such delays—this time around. I honestly believe they don’t want to go through this process again if they can avoid it.

Trying to power-lift everything—the development tools, SQL Server, the client and server operating systems—to the next level has proven a serious challenge even for the most mighty software development organization on the planet. WinFS—once the third leg of the Longhorn stool—is gone for now. And Allchin readily admits that Microsoft has even more issues it must address if it is to fully achieve its vision for Longhorn—issues that don’t have obvious answers. (One feature, for instance, would make it easy for users to suck everything—OS, applications, data and all—from an old machine to a new one. But how does Microsoft deal with the licensing issues for software keyed to a single machine? Allchin doesn’t know yet.)

The fact that Windows XP is actually pretty solid right now will only work against Microsoft if Longhorn doesn’t end up being something really special. (Why upgrade if you’re quite comfortable with what you have?) Botch something this complicated, and Microsoft could spend the rest of its days as little more than a support organization dedicated more to issuing patches than to advancing its technology agenda.

Twenty years of Windows. Impressive. But the next couple of years will determine if the timeline continues for another decade.